-
How to pick up speed in Radiology?
How many studies (cross sectional/Radiographs/US/NM) you should read in average per hour or per shift? How long should reading a study take?
what if you put quality reports but youre really slow compared to your peers, how to pick up speed?
Thank you -
77 Replies
-
Way too many variables. Look look at the average RVU generation other people make on that same shift.
-
Many many many factors
What level are you? Most important, esp initially
How many times are you scrolling through the exam?
Do you have “wasted” scrolls (going to the top or bottom without looking at anything)?
How much are you looking stuff up
How many times / how much time are you spending do determine if something is normal / abnormal (again, experience based)?
Are you looking at the dictation while you are dictating, or editing at the end (one is MUCH faster)? related: Do you have your search patterns memorized?
Are you using window level presets? (usually the number keys on most PACS)
Have you trained your VR approriately?
How many interruptions do you have?etc
etc
etc
-
Is it faster to proofread the dictation as you go or after you are through?
-
You can go much faster if your search routine is shorter and high yield, you only look once at everything quickly, you let a lot of clinically unimportant things go, you sign off immediately after going through the search, keep reports short, limit comparisons to the last one only in most cases.
Someone who is wonderfully accurate and detailed with low output will suffer much more than the average guy with average output in many practice settings. Its better to keep your job then to impress colleagues and referrers with your radiology skills. If you can go fast and be accurate then you are lucky,
-
Minimize distractions: do not stop your search pattern or dictation because you get a notification on your phone or to check your email or because you are trying to carry on an interesting conversation with the fellow or attending. That should be obvious but has to be said. Don’t take a break to do the above in between cases until you get through a certain number of cases, e.g. 5 or 10. When you settle into a particular practice and know how fast you need to work, you can then be like me and get through 6 episodes of Bosch in regular workday. My rule is to never watch movies or TV shows that I’ve never seen before, it’s too distracting to follow the plot line, but re-watching the good ones allows me to ignore it when I need concentrate hard and then pick up the plot line again when it goes back to routine. This is an advanced tip, not for trainees.
Get through the normals faster. Complex cases require attention and time.
Be less descriptive about things that don’t matter. Don’t describe every benign cyst in the kidney in three dimensions with image numbers. “Multiple benign cysts seen in both kidneys, largest is 2.5cm in the right kidney” is enough.
Be more interpretive and less descriptive. Not “T2-hyperintense rounded lesion measuring A x B x C with progressive peripheral enhancement and centripetal fill-in on delayed phase, seen on image number…” Goddangit, you are not paid by the word. “2.5cm benign hemangioma in segment 5 is unchanged.”
If you are not wedded to structured reporting… yeah I know, it’s better… Go through your search pattern and only dictate everything incidental first and skip the bad stuff, i.e. that inflammatory mess in the right lower quadrant. Then go back and look at the ‘main issue’ in detail, in different reformats or sequences, compare to priors etc, all the while thinking about ‘WTF is this thing and what will I say?’ As soon as your thoughts are gathered, describe it and any pertinent negatives related to the ddx of what the thing is. Then as a final sentence in your findings section, run though the number structures (thinking of your search pattern) “The A, B, C, D etc are normal.”
Impression section doesn’t have to mention every flippin osteophyte and benign cyst and empty sella syndrome. Just what’s important and relevant, hopefully just a few impression numbers.
If you don’t know what’s going on, realize that staring at it for another 2 minutes will rarely generate additional genius revelation. Most of what you’re going to see and say will be seen and said in a first few minutes, then diminishing returns. Someone has once said, “If you’re going get it wrong, get it wrong fast!”
The greater the complexity of findings, the shorter your report. Clinicians want you to show them the big picture, they want to be shown the forest, not the trees. Don’t regale them with a litany of Modic this and Chilaiditi’s that and the grand disease of Scheuermann. The report may already be long and they are in no mood to be screwed around with. If a surgeon is having his fine dining experience with trophy girlfriend interrupted (dinus interruptus?) to come into the hospital and evaluate an ER patient, forcing him to then ask an R2 about this “Chilaiditi’s phenomenon” may just get you a, “Are fu**ing kidding me?”
If you get dictation paralysis, then dictate notes and impression for yourself to sign off “later.” The idea that you think of what you’re dictating as not final but rather something you’re going to curbside another radiologist later…. and that may be enough to loosen the lips and get things flowing. Sometimes you just gotta talk and say something and you look back at the end and think, “okay, not my best work, but hey whatever, next case.” Not every case has to be your masterpiece.
Say you develop a rhythm and a momentum in your work and you are on fire. Don’t let a stray weirdo outpatient case disrupt you. “Whoa… what is this weird stuff?” Assuming it’s not acute, decide right away that you are going to SLOW COOK this one: you flip through the images and take it all in, pause for a second to see what your initial impression , then draft it and move on. Over the next hour, while you dictate other cases, there is a part of your brain that will keep chewing on this case until you are comfortable with what you’re going to say. This is the same part of your brain that tells you (‘oh sh*t!’) as you are about to fall asleep that you forgot to include a significant incidental finding in the impression section of the one dictation. When you go back to dictate that case, you’ll know what to say.
That’s all that comes to mind.-
Dont have much to add to flounce. Just wanted to say I love bosch. Great show.
-
It depends. If you don’t know it and it’s not stat, then draft it and come back. Or if you know it and you know it’ll be a time suck (met follow up with a maybe progression/complex post op, etc), then draft it and come back.
If I’m behind, I’ll do all my cross sectional stuff first then all the radiographs. I set stuff to auto load and recheck the list every 20 minutes for a stat. Headphones set to shuffle, iPhone put face down, and just dictate until it’s done.
I’m pretty ADD, so it took some effort to realize my inefficiencies.-
Lot’s of great advice from Flounce.
Here are a few more:
–Dictate/speak as you look in your search pattern. The process of generating the report and looking at the images should happen at the same time. Almost all the “slow” readers I’ve ever known will look and think first, then dictate second. 95% of the cases you can easily dictate as you simultaneously look. Once or twice an hour you may run across something difficult/strange which will require more thinking first.
–Know your audience. When I finished training and started my PP job, I wanted to show off all my knowledge and intelligence with very detailed words and classification systems, especially within my fellowship. The overwhelming majority of referrers in the community (ER, IM, FP and especially NPs) do not understand and are frankly extremely confused for me to say something like Schatzker IV fracture. It wasted time with the extra words to dictate/describe it and much more so with the inevitable phone call from them to explain it.
–Invest the time every 3-6 months to tweak your voice recognition templates. Saving even just a few seconds on every study will add up over the course of a day and week.-
Half of what you learn during training to put in your dictations are mental masturbations.
Reports from academic centers are too wordy.
Being a radiologist in the community practice is way different than a radiologist in tertiary academic center.-
I wouldnt toss the word benign around so much though. Less is more, let them go the extra mile after you if they wanna hear the word benign. Its anyhow on them if they do something dumb for a hemangioma or cyst, not you. I actually saw a rad lose a case once and get completed hosed over something they called benign when apparently it was too small/study wasnt Good enough to know for sure it was benign – personally I think it was bs and was an obviously benign thing but they harped on that word and used it as a A was said therefore B didnt happen chain of nonsense I wont get into.
Bottom line less is more -
It’s an interesting topic.
My take. If you say “there is a simple cyst in the left kidney measuring 2.5cm” or “there is a 2.5cm hemangioma in the liver,” there are many clinicans who don’t know that you are saying it’s benign and that it doesn’t need to be followed up. So they are in clinic looking at the screen and wondering what to do with it, sometimes they will message a nephrologist to advise if follow up is necessary, sometimes they might pick up the phone to ask a radiologist. Point is that they it takes them extra time.
I like to add the word benign, and if the ordering provider is particularly dense, occasionally I’ll even add “no follow up needed.” Yeah it takes extra words and in some cases, increases liability, but so be it. We want our reports to be helpful and if I know a clinician is going to read something and think, “what does that mean?” or “what am I supposed to correlate it with?” then I’d rather clarify it rather than make them call me or one of my partners up for additional consultation.
This is just a personal philosophy, and my clinicians like it. I try not to dance around. I don’t know that there is any liability in calling it benign vs. calling it a “non-enhancing T2 hyperintense circumscribed oval collection in the kidney consistent with simple fluid.” If you miss a cancer, you missed it, and if it’s benign, you’ve done your clinician a service by being straightforward. But even if there is some liability in offering my interpretation of the images – that I see it as benign – I guess I am okay with a little liability to make most of my reports more helpful rather than play CYA on things that are obvious. It’s why we are paid the big bucks, so to speak. Just my 2 cents. -
I get what youre saying. If I gave more details in what happened I think you would be enraged like I was. Its possible it was one of those no matter what happened they were gonna find a way to pin the rad cases but its sometimes amazing how creative lawyers can be.
-
Having said that, Ive never in 10 years received a call or ever noticed follow up for what was described as a simple cyst, even though the word benign wasnt used. Not saying it doesnt happen, but thats my experience.
I have I think had 1 or 2 times where a hemangioma was followed that didnt need to be, so I get you there. I dont count a large hemangioma with ruq symptoms as an unnecesary followup personally) -
I think your providers are smarter than mine. Maybe we have more nurse practitioners. We get follow up studies for ovarian cysts (that don’t require it), small thyroid cysts, liver cysts, renal cysts, hemangiomas, and all sorts of things that should be obviously benign. If my experience was yours, I wouldn’t have to say it was benign, either.
-
I use “without suspicious features.” Typically for nodes or thyroid nodules found on a c-spine or chest CT.
-
To be fair, how does a non-radiologist know that a cyst in the liver or kidney is probably benign, but a cyst in the pancreas needs follow-up per the ACR guidelines? More helpful to not assume standard radiologist knowledge and be clear. When I see an ovarian cyst, I specify “within normal limits of appearance for age.” When I see a pancreatic cyst, I give the specific recommendation per the ACR white paper.
-
Knowing what should be followed up isnt just standard radiologist knowledge, its standard doctor knowledge. If the noctors dont understand, then thats on them.
-
There are many family practice and internist physicians who don’t know about the Fleischner Society recs for lung nodules or ACR white paper recs for pancreatic cysts, while standard knowledge for rads. Why make things unnecessarily harder for them in terms of work compared to knowledge we have at our fingertips? Sounds like a whole lot of inefficiency to me.
-
to OP:
1.You can become faster if you know more rad path correlation as it boosts your confidence and you don’t dither with multiple differentials and make yourself and everyone else confused.
2. If you can convey the information with less words, you will dictate it faster. A lot of words and fluff is just noise and will slow you down. -
Short reports.
-
Many good suggestions on becoming more efficient.
Bottom line, there is a point where you cannot go faster without cutting corners. It takes a finite amount of time to get through a checklist, in your head or on paper, and that’s that. When you reach that point, you reach the asymptote.
Daniel Kahnemann’s book “Think Fast, Think Slow” nicely addresses the issue. We all utilize heuristics as shortcuts, which give us the illusion we are making more efficient decisions, but we are not. There are all sorts of biases through which we deceive ourselves in the course decision making.
Ultimately, you still have to look at the study, compare with old exams [which many skip at their peril] and correlate with the medical record [which is under-appreciated in value.]
You will get to a point where you will not get any faster; it’s just not physically possible, you have to learn to live with that.
If you are in a practice that doesn’t get it; find a new job. -
Dergon 1998: “In the central metaphysis of the distal right femur there is a 1.3 by 1.4 by 2.1 cm lesion. It has a lobular morphology on its periphery with predominantly increased signal on T2 weighted images, decreased signal on T1 weighted images, and small specked areas of markedly deceased signal on both T1 and T2 weighted images corresponding to arc and whorl mineralization seen on the previous radiograph. The lesion shows no significant endosteal scalloping, has no cortical breakthrough and no soft tissue mass. There is no adjacent soft tissue edema.”
Dergon 2020 (and since about 2005): “Small distal femoral enchondroma.” -
Limiting ocd tendencies will be high yield for some.
-
Thank you so much for the feedback!
-
I use ” the cyst (or whatever) demonstrates benign imaging characteristics”
Says it all. -
Some absolute GOLD from Flouce in that post.
-
Ya if you get down to the truth of it, a big part of the speed of some folks is not bothering to compare priors(or spend the time to get priors that arent immediately available), or checking medical record for relevant info. That last hump to get past when becoming a fast reader is rarely one where you come out on the other end a better doctor
-
Same here, only I try to add, if relevant, stable since comparisons from 35 BC.
-
Quote from dergon
Dergon 1998: “In the central metaphysis of the distal right femur there is a 1.3 by 1.4 by 2.1 cm lesion. It has a lobular morphology on its periphery with predominantly increased signal on T2 weighted images, decreased signal on T1 weighted images, and small specked areas of markedly deceased signal on both T1 and T2 weighted images corresponding to arc and whorl mineralization seen on the previous radiograph. The lesion shows no significant endosteal scalloping, has no cortical breakthrough and no soft tissue mass. There is no adjacent soft tissue edema.”
Dergon 2020 (and since about 2005): “Small distal femoral enchondroma.”
So from an academic dictation to a community rad dictation?
-
Cross sectional 1-2 minutes drad123- do these super fast rads even have a search pattern or is it a gestalt type read where they scroll and if nothing hits them in the eyes its normal?
-
Quote from Drrad123
Cross sectional 1-2 minutes drad123- do these super fast rads even have a search pattern or is it a gestalt type read where they scroll and if nothing hits them in the eyes its normal?
I don’t know for sure but some appear to see everything quickly. I take longer.
It’s an art form. Minimalism.
-
I dont see any art form in reading a belly ct. its mechanical. Go through search pattern, find abnormality, describe, diagnose. Rinse repeat.
You can limit search patterns, gestalt it without scrutinizing individual structures hope abnormality hits you in the eye, dont compare to prior, limit words. These will speed you up at your own risk.
-
Quote from Drrad123
I dont see any art form in reading a belly ct. its mechanical. Go through search pattern, find abnormality, describe, diagnose. Rinse repeat.
You can limit search patterns, gestalt it without scrutinizing individual structures hope abnormality hits you in the eye, dont compare to prior, limit words. These will speed you up at your own risk.
How many cases do you read per hour mixed modality? I observe and admire the fast readers who do it well. Doing it well means comparing.
-
Radiograph reports that are only a couple of lines don’t deserve separate Findings and Impression sections. I’ll only add an impression line to radiographs if report is overly complicated, everything can be neatly summed up with one word or phrase: e.g. CHF or if I am recommending a followup study like CT for further eval just to make sure it does not go missed. This def saves some time when plowing through large list of radiographs compared to how we dictated these cases in residency.
-
Quote from MD20/20
Radiograph reports that are only a couple of lines don’t deserve separate Findings and Impression sections. I’ll only add an impression line to radiographs if report is overly complicated, everything can be neatly summed up with one word or phrase: e.g. CHF or if I am recommending a followup study like CT for further eval just to make sure it does not go missed. This def saves some time when plowing through large list of radiographs compared to how we dictated these cases in residency.
Yes it does. I have seen rads fly through 50-60 cases of portables and overnight ER plain films in an hour doing this. -
I can easily read cts from the ER in 2-3 minutes. You need the right voice recognition too. And for the young ones out there, that Dergon paragraph is key. Once you leave academia, not only do most of your referrers not need the detailed report, a lot of them will be annoyed at you for it. You have to know your audience. It comes with time.
-
PET/CT
I have seen rads turn this into a lucrative study.
They don’t break out anatomic section like head neck, chest, abd pelvis
Limit CT findings to major or clinically relevant
Dictations times under two minutes unless complicated. -
The best way to get faster is to know what’s important.
The difference between an expert and a novice is that the expert focuses on the few key findings. The novice looks at everything equally. -
I have worked places where my reports where too short and other places where they were too long.
-
Quote from RadiologyTutor
The best way to get faster is to know what’s important.
The difference between an expert and a novice is that the expert focuses on the few key findings. The novice looks at everything equally.
+1 -
90% of our neuroticism when dictating a report is because we care about what our radiologist colleagues think of us, not what the patient outcome will be or what the referring clinician will think. That’s why you mention the focal fatty sparing adjacent to the falciform ligament, the Nabothian cysts, and the small vertebral body hemangiomas.
So find a way to lose some respect for your colleagues, and you will stop caring about their opinions and will read much faster. -
Quote from Dumb Luck
90% of our neuroticism when dictating a report is because we care about what our radiologist colleagues think of us, not what the patient outcome will be or what the referring clinician will think. That’s why you mention the focal fatty sparing adjacent to the falciform ligament, the Nabothian cysts, and the small vertebral body hemangiomas.
So find a way to lose some respect for your colleagues, and you will stop caring about their opinions and will read much faster.
+1
Hear! hear!
Though I think you mean to lose respect for what your colleagues think about you, not to lose respect for you colleagues. I could be wrong. You can love your colleagues but not care much what they think of you, so long as you know your are not doing anything wrong.
There is a parable of life in this. Sometimes you can become a better radiologist by caring more about your radiologist colleagues but caring less about what they think of you. Sometimes you can become a better human being by caring more about your fellow humans but caring less about what they think of you. -
Removed due to GDPR request
-
I definitely mention a 2.3cm benign hemangioma in hepatic segment 5 so that I won’t get that call, I don’t blow it off. But I don’t say it’s a T1 hypointense, T2 hyperintense, no drop out on In- and Out- phase imagin, no reduced diffusion, has progressive centripetal and ultimately complete enhancement, measuring 23mm x 21mm x 19mm with circumscribed but acceptably microlobulated margin, and representative images are series 2 image 13, series 8 image 10, and series 12 image 6.
-
Quote from 67ED5CC042435
You guys never get calls because the ordering doc was looking at the study themselves and want to know why we didn’t describe a mass in the liver (e.g. fatty sparing) and when you say it’s nothing they press for an addendum? I dictate the way I do at this point to minimize the number of unnecessary phone calls.
I will occasionally get calls for re-review. (Usually an orthopd wanting an equivocal or non-finding described as pathologic so they have an excuse to do surgery)
I will only addend if I agree.
But I’m not going to get into the ethical gray-space of up-scaling my reporting of the pathology in order to make referrers happy.
I *do* try to avoid phone calls by making sure not to use words that freak out primary care docs. Long ago “spina bifida occulta” was changed to “normal variant posterior element fusion defect” -
Detailed reporting is more of a subspecialized thing. The rad specialists try to “prove” their value with longer reports.
I have had more issues with other rads reading my reports and saying they weren’t detailed enough than with referring docs. -
Quote from Flounce
It’s an interesting topic.
My take. If you say “there is a simple cyst in the left kidney measuring 2.5cm” or “there is a 2.5cm hemangioma in the liver,” there are many clinicans who don’t know that you are saying it’s benign and that it doesn’t need to be followed up. So they are in clinic looking at the screen and wondering what to do with it, sometimes they will message a nephrologist to advise if follow up is necessary, sometimes they might pick up the phone to ask a radiologist. Point is that they it takes them extra time.
I like to add the word benign, and if the ordering provider is particularly dense, occasionally I’ll even add “no follow up needed.” Yeah it takes extra words and in some cases, increases liability, but so be it. We want our reports to be helpful and if I know a clinician is going to read something and think, “what does that mean?” or “what am I supposed to correlate it with?” then I’d rather clarify it rather than make them call me or one of my partners up for additional consultation.
This is just a personal philosophy, and my clinicians like it. I try not to dance around. I don’t know that there is any liability in calling it benign vs. calling it a “non-enhancing T2 hyperintense circumscribed oval collection in the kidney consistent with simple fluid.” If you miss a cancer, you missed it, and if it’s benign, you’ve done your clinician a service by being straightforward. But even if there is some liability in offering my interpretation of the images – that I see it as benign – I guess I am okay with a little liability to make most of my reports more helpful rather than play CYA on things that are obvious. It’s why we are paid the big bucks, so to speak. Just my 2 cents.
I’m not sure why people think that NOT calling something benign somehow protects them. By that logic, by not recommending a follow up, are you also liable; ie. If you’re the clinician and the rad DIDN’T tell you to follow it up, isn’t it safe to assume it’s benign?
Point is, if it’s malignant or potentially malignant, the responsibility is on both of you to follow it up. Same goes the other way. When findings meet the criteria for benignity, it is safe to say so. -
We have a role to play; to be useful. Being cryptic about findings provides an illusion of protection. It’s embarrassing to read some of the CYA reports. I get calls from clinicians asking what does he/she mean? And I say, I’m not sure, you are going to have to call them; and then give my personal opinion.
I use “benign appearing findings” as a term all the time.
Many older clinicians are more confident in dealing with imaging, or not; younger ones, esp extenders, want their hands held and recommendations spelled out to ridiculous degrees.
I have one partner who is very good, and will say benign findings, but then follows up almost everything anyway. Makes it easy as the next rad, can almost guarantee you can say no change of benign appearing lesion.
It’s got to the point where some patients are on a conveyer belt of followup in perpetuity.
We all hedge, sometimes it is impossible to avoid. But minimizing it should be a goal. The conveyor belt has to stop somewhere.
When you know the physicians, and see the same patients over time, it is easier to do. As a teleradiologist interpreting from who knows where, I can see how it would be more difficult.
-
Quote from boomer
We have a role to play; to be useful. Being cryptic about findings provides an illusion of protection. It’s embarrassing to read some of the CYA reports.
+1000
Drives me crazy! I stress it over and over again to my residents.
Somehow, some radiologists have become convinced that hedging is medico-legally protective.
I tell my residents that whether your impression reads “No radiographic evidence of definitive acute pulmonary process on this limited AP chest” or “Normal Chest” you are still going down if you miss a 2 cm pulmonary nodule. …
so you might as well pick the style that your referrers find useful and that doesn’t make you (and by association all radiologists) look like an obfuscating douche. -
Best way to pick up Speed is to look for the shadiest radiologist or staff member and ask him/her where to score drugs.
-
Another good way is to stop caring much and lose your conscience. Ive seen some do that and they are some of the fastest.
-
Quote from Drrad123
Another good way is to stop caring much and lose your conscience. Ive seen some do that and they are some of the fastest.
You mean not being obsessive? -
Quote from dergon
Drives me crazy! I stress it over and over again to my residents.
Somehow, some radiologists have become convinced that hedging is medico-legally protective.
I tell my residents that whether your impression reads “No radiographic evidence of definitive acute pulmonary process on this limited AP chest” or “Normal Chest” you are still going down if you miss a 2 cm pulmonary nodule. …
so you might as well pick the style that your referrers find useful and that doesn’t make you (and by association all radiologists) look like an obfuscating douche.
I agree with you for the most part.However, if there is some DJD in the thoracic spine, technically you can not call it normal chest. That’s the reason most people say “No evidence of acute process”.
Similarly in CT head if there is some white matter disease, you can not call it “normal exam”. -
Quote from Hospital-Rad
Quote from dergon
Drives me crazy! I stress it over and over again to my residents.
Somehow, some radiologists have become convinced that hedging is medico-legally protective.
I tell my residents that whether your impression reads “No radiographic evidence of definitive acute pulmonary process on this limited AP chest” or “Normal Chest” you are still going down if you miss a 2 cm pulmonary nodule. …
so you might as well pick the style that your referrers find useful and that doesn’t make you (and by association all radiologists) look like an obfuscating douche.
I agree with you for the most part.
However, if there is some DJD in the thoracic spine, technically you can not call it normal chest. That’s the reason most people say “No evidence of acute process”.
Similarly in CT head if there is some white matter disease, you can not call it “normal exam”.
Can call it “negative.”
definition: characterized by the absence of distinguishing features.
Normal is ok too.
definition: usual, average, or typical state or condition.
So mild djd in a 60 y/o CXR can still be called “normal,: as it is a typical condition for that age.
Normal does not mean without abnormality, but rather an expected condition for that demographic. -
Quote from boomer
Can call it “negative.”
definition: characterized by the absence of distinguishing features.Normal is ok too.
definition: usual, average, or typical state or condition.
So mild djd in a 60 y/o CXR can still be called “normal,: as it is a typical condition for that age.Normal does not mean without abnormality, but rather an expected condition for that demographic.
What is the expected amount of atrophy or white matter disease for a 65 year old man?
-
I like within normal limits. And use liberally. Says normal and negative but with enough wiggle room to gloss over those DJD changes and other non significant things.
-
I use “unremarkable findings” a lot.
-
I have seen lung screenings reports
Findings: no lung nodules are seen.
Impression
No lung nodules
Lung rads 1If theres a module.
Finding: There is a 5 mm nodule in right lung.
Of course they are one of the fastest readers in group
-
I tell my residents that whether your impression reads “No radiographic evidence of definitive acute pulmonary process on this limited AP chest” or “Normal Chest” you are still going down if you miss a 2 cm pulmonary nodule. …
so you might as well pick the style that your referrers find useful and that doesn’t make you (and by association all radiologists) look like an obfuscating douche.
Who are those referrers? For our hospital based practices a majority of them are PAs or NPs. Are we supposed to assume that all of them understand that chest X-rays are very limited for everything?
What about the patient with chest pain that reads her normal CXR report to mean she can rest easy and ignore the chest pain because she has normal imaging? Turns out she had a 8 cm nodule invisible by CXR but which would have been present on CT and later making an appease as a light bulb bright 2 cm nodule on PET-CT.
That may seem paranoid, but Ive encountered situations like this and Ive learned to never underestimate the presence of ignorance in and around healthcare. Thats why I like to state things within the limits of studies and let those reading it to know so.
Sincerely,
Obfuscating Douche, MD -
Don’t duplicate statements from findings in impression.
-
Getting rid of the impression on plain films. I have seen this done with some high volume rads in certain areas.
-
Limit descriptions, Limit negatives.
I have seen super high volume rads dictate x rays in under 30 seconds and cross sectional in under 1 min unless complicated then under 2 min.
-
I learned right out of fellowship from my previous partners. My reports were [i]detailed[/i]. I touched on every major organ in every exam. (ie, liver, GB, pancreas, adrenals, kidneys, bowel, vessels, pelvis, spine). Their reports to my amazement were ludicrously short (ie, “there is an obstructing 4mm calculus in the distal third of the right ureter with mild hydro. No other abnormal findings”. But they didn’t [i]miss [/i]stuff. To my further amazement, the clinicians preferred this. Fast Fwd to now, I still feel that the report should touch on pertinent negatives, so I use templates. But for things that are ubiquitous like simple cysts in the kidneys, phleboliths in the pelvis, old schmorls nodes in the spine, etc, I spend zero time. I may not even mention them at all, because ubiquitous things by definition are exceedingly common. And presuming you are newer, after you read tens of thousands of a particular exam, you can scroll quite quickly through and you “automatically” recognize something is off, because you have seen oodles and oodles of [i]normal[/i] and your subconscious takes over to tell you something doesn’t quite fit. Speaking of normals, if you spend a few minutes creating templates, negative chest x-rays, msk x-rays, head CTs, etc reports are a simple click. Just don’t say things like the sinuses are clear, because that is too variable. And if you have it in your template, make sure that is in your normal search pattern for that exam. I have seen plenty of people burned by their overly wordy templates. Trying to sound like a smartypants isn’t in your best interest.
-
Synchronous reading and dictating. Hard to do with voice recognition and templates but easy with transcriptionists.
-
I’m reading these caricature reports and laughing. Lots of helpful tips.
I think what Flounce says about caring more for your fellow human but less about what they think of you rings true to not just making you a better human but making you just more comfortable with everything in life. Not easy to practice but a good ideal to strive for.
I like the statement about our neuroticism when dictating studies. It feels like many times people dictate with two (or even more) separate audiences, one always being another radiologist. I’ve been increasingly dropping clinically insignificant incidentals, I think I’ll continue experimenting with this perspective in mind.
I wonder though, at my institution, faculty are reading 10-15 cross sectional studies a day (in a setting without many distractions or other duties), sometimes less, and rarely more. Is this normal for an academic setting? What’s the longest time a radiologist should realistically spend for a complicated study?
At the same time, when residents are reading multiple folds that number across multiple subspecialties in a fatigued state overnight while answering dozens of phone calls and having to protocol studies, they get reamed for missing or misreporting things (which many times turn out to be clinically inconsequential). The feedback is often snarky or intentionally humiliating. So then feelings of bitterness and resentment naturally build. -
Quote from irfellowship2020
I wonder though, at my institution, faculty are reading 10-15 cross sectional studies a day (in a setting without many distractions or other duties), sometimes less, and rarely more. Is this normal for an academic setting? What’s the longest time a radiologist should realistically spend for a complicated study?
10-15 cross sectional studies per day?!?!?!?! Is this even on planet earth? My fellowship was at a major academic center and as fellows would average around 30 complex MRI per day, even more if CT. Attendings would probably over-read closer to 50-80 cross sectional cases. In residency I did a hybrid academic/private and volume was probably similar if not even higher. Do your attendings make less than 200k or is this the middle of nowhere?? 10-15 studies per day is astonishingly low. Expect to read 5-10x this volume once you’re out in the real world.
-
Great advice on this thread, I want to ask a slightly different question. Currently starting my last year of residency, so have ~2 years before private practice. My speed has gotten a lot better and I can handle our call without getting too far behind, but realizing pp will be significant more volume. Is there anything you wish you would have done differently in the last years of training to become faster for private practice, other than just generally seeing more cases and becoming familiar with a wide range of pathology? I’ve switched to a mouse with 12 side buttons (24 with shift) which takes care of all my windows and annotations, so I don’t have to touch the keyboard. I’ve also realized I scrolled to fast and tried to slow down but take fewer passes through the study which seems to help. Wondering if anyone has tips, or other bad habits they might have had to get rid of after completing training.
-
1. Practice getting through the normals faster. (Normal doesnt mean no findings, it means no findings relevant to the clinical question but often with many incidentals. Incidentals are like quicksand, dont get bogged down.)
Most of what you read in the community is essentially normal, so getting faster through those – alone – will make you much faster.
2. Spend more time looking at images and thinking; and less time dictating/fussing over the report. Impression #1 needs to address the clinical question, using clinician language, not radiologist language.
-
Quote from InsertNormalChest
Great advice on this thread, I want to ask a slightly different question. Currently starting my last year of residency, so have ~2 years before private practice. My speed has gotten a lot better and I can handle our call without getting too far behind, but realizing pp will be significant more volume. Is there anything you wish you would have done differently in the last years of training to become faster for private practice, other than just generally seeing more cases and becoming familiar with a wide range of pathology? I’ve switched to a mouse with 12 side buttons (24 with shift) which takes care of all my windows and annotations, so I don’t have to touch the keyboard. I’ve also realized I scrolled to fast and tried to slow down but take fewer passes through the study which seems to help. Wondering if anyone has tips, or other bad habits they might have had to get rid of after completing training.
Good that you are asking these questions now.
Volume, variety, and independence. Moonlighting if available to you. Forces you to make quick decisions.
Developing your own internal search patterns. Develop templates for reporting cases. Make sure to get copies of templates to take with you where ever you end up. Review others templates and look for opportunities to make more efficient.
-
Take independent call.
Follow your cases since you have more free time as a resident
See as much as you can. Be a fly on the wall when your attending is discussing something interesting with your colleague.
-
Try to learn as much relevant stuff as possible. To to learn how to stage tumors, grade MSK injuries or vascular findings, form a differential from a finding etc without having to look up something. This of course, isn’t required. You can just learn as you go as an attending and the speed will come that way also. But, getting head start is key.
Learn what clinicians want to know, so you always have it available somewhere in your report. This is done by going to multidisciplinary clinics or seeing what a good attending says in their reports.
I’m a macro guy. Can have blurbs on what to do with incidental findings. Or just say a common sentence. Saves proofreading time.
Make sure your powerscribe or other dictation software is adequately trained. Speak clearly to the mic. This way, you will make less errors. If you speak too quickly or not at the correct level, there may be lots of VR errors. You will still have to proofread. I swear, I say the correct side, but the opposite side shows up on the report. Hypoinflated becomes hyperinflated or vice versa. Ascending comes out as descending. The ordering docs may think you are sloppy if there are frequent errors in the report. -
Still fairly early in my career but here are some observations/tips:
-The most efficient (fast and accurate) readers that I have come across have succinct, short reports. They save time on dictation length/content rather than cutting corners with looking at images/search pattern. They also know the precise clinical question (not necessarily congruent with given hx).
-Practice efficiency with ICU CXRs. Very common to see Findings/Impression combined stating no significant change. Call/ER shifts also help with internal pacing
-Know your audience. Are succinct, short reports sufficient for the referrers and patients that actually get to view them? Are you reading for subspecialist docs or mid-levels?
-Know your practice. Are you in high volume or average volume practice? I’ve been in both. Average volume allows me to actually look stuff up and come up with a coherent differential etc. Not so much in high volume.
-Know when thorough comparison/correlation to prior studies (including ones not directly related) is needed.
-Understand how your personality/temperament influences your dictation style. If you’re an OCD type person, you need to set limits on time/case. Understand that despite whatever you do, you will miss stuff and receive criticism, much more so than receiving praise
-Be realistic. For most rads, it takes some time and experience to get fully up to speed. Knocking out 17-18K RVU in year 1 is the exception not the norm
-
I think those that have the easiest transition to attending volumes moonlighted during fellowship year. I honestly think this should be done by all but not always available and not easy if in an ACGME fellowship but that experience is priceless.
-
Sometimes I just want to send them a two-liner rather than having to go through the motions:
FINDINGS/IMPRESSION: Your patient is dying from progressive lung cancer. He didn’t have a PE on Monday and he doesn’t have one now, he has lung cancer, that’s why he can’t breathe. -
One of the pragmatic neurooncs sees their patients immediately after scans and on bad ones will email us to let us know they’re sending the patient hospice so we can keep impressions down to “Significant progression, worsening shift” instead of measuring each of the 10 lesions a rad onc zapped
-
-
-
-
-
-
-
Quote from api7342
Is it faster to proofread the dictation as you go or after you are through?
After
-
Maybe I’m saying this because I was always a fast reader, but I would say focus on learning during Residency. Spend time going over normals/pathology, googling stuff, and asking questions of your attendings and seniors. Read books and articles.
Most of all, enjoy your free time and work — these are some of the best years you will ever live. You have the rest of your career to optimize performance on the hamster wheel.
-
-
-
-
